Durability of concrete and cement composites

Spatial distribution of pores in concretes: the ITZ

Pores, especially the larger pores visible in backscatter SEM, are not distributed entirely uniformly within cement paste in concrete. One commonly discussed aspect of this non-uniformity relates to the interfacial transition zone or ITZ; i.e. the local region of cement paste surrounding (and in contact with) sand and coarse aggregate grains. Cement paste in concrete has been looked on by many authors as consisting of two distinct entities: `bulk’ cement paste removed from the local influence of sand or aggregates, and `ITZ cement paste’ close enough to the nearest aggre- gate to be affected by its presence. `Close enough’ has been frequently redefined, but it is generally taken as within 35 um of the nearest aggregate. While various differences between the two `cement pastes’ can be detected by image analysis ± especially in the contents of residual unhydrated cement ± the important difference usually is considered to be the pore content, specifically the content of pores large enough to be tallied in backscatter SEM. Scrivener41 reported spatially-related backscatter SEMimage analysis porosity data derived from Crumbie,42 for a w:c 0.40 concrete evaluated at several ages. The average detectable pore content found in `bulk’ paste in the mature concrete was ca. 10%. Average pore contents reported for successive ITZ `slices’ taken progressively inward from the bulk paste to the aggregate interface itself were progressively larger as the aggregate was approached, and jumped remarkably for the innermost 5 um slices in the data reported by Scrivener41 no direct comparison for the innermost 5um slice was possible. However, extensive visual examinations of the areas immediately adjacent to the aggregates reported by Diamond and Huang43 for their specimens showed only a limited content of detectable pores, and were not at all commensurate with the very high values reported by Scrivener.41 Indeed, it was found that a significant portion of the areas immediately bordering aggregates were entirely blocked by Ca(OH)2 layers deposited directly on the aggregate surface, and commonly extending 5 um or more into the surrounding paste. Elsharief et al.44 recently reported average pore contents for a w:c 0.40 mortar (not a concrete) to be about 8% at maturity (180 days); the average pore content found for the innermost segments adjacent to the aggregates was ca.15% at maturity, higher than those reported by Diamond and Huang,43 but very much lower than those reported by Scrivener.41 Diamond and Huang43 called specific attention to the fact that whatever the average values, compilation of averages masked large variations in porosity among adjacent sampling units. Some adjacent segments at a given distance from the aggregate were highly porous; others were almost lacking in detectable pores. The general inhomogeneity of the interfacial zone ITZ has also been stressed by Scrivener41 and others.
The possible effects of the existence of a more porous ITZ on various proper- ties of concrete have been explored by a number of authors. A representative result was that of Delagrave et al.,45 who found that the numerous ITZs in the mortar they studied had no measurable effect on the kinetics of leaching of calcium hydroxide. Many studies aimed at elucidating the effects of ITZs on permeation capacity-related properties similarly showed little or no effect. Summarizing some of them, Scrivener41 concluded that, while the higher porosity in the ITZ might be expected to increase permeability, the presence of impermeable aggregate particles, around which the ITZs form, combined with the lower local w:c induced in the `bulk’ paste will work in the opposite sense; thus the increased porosity of the ITZ is of minor importance compared to other factors.

Related Articles